Deep Learning Prediction of TERT Promoter Mutation Status in Thyroid Cancer Using Histologic Images
Abstract
:1. Introduction
2. Materials and Methods
2.1. Study Population
2.2. Histologic Evaluation
2.3. Dataset Preparation
2.3.1. Annotation of Tumor and TERT Positives
2.3.2. Downsampling Ratio
2.4. Whole Architecture for TERT Prediction
2.5. Data Split
2.6. Classification of Tumor Areas Using CNN
2.6.1. Patch Filtering
2.6.2. Color Transformation as Image Preprocessing
2.6.3. CNN Model Training
2.7. Prediction of TERT Promoter Mutation Status Using CRNN
3. Results
3.1. Patient Cohort
3.2. Tumor Classification
3.3. TERT Classification Performance Results Using the CRNN Model
3.4. Whole Inference Process
4. Discussion
5. Conclusions
Supplementary Materials
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- Sung, H.; Ferlay, J.; Siegel, R.L.; Laversanne, M.; Soerjomataram, I.; Jemal, A.; Bray, F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J. Clin. 2021, 71, 209–249. [Google Scholar] [CrossRef] [PubMed]
- Milano, A.F. Thyroid Cancer: 20-Year Comparative Mortality and Survival Analysis of Six Thyroid Cancer Histologic Subtypes by Age, Sex, Race, Stage, Cohort Entry Time-Period and Disease Duration (SEER*Stat 8.3.2) A Systematic Review of 145,457 Cases for Diagnosis Years 1993–2013. J. Insur. Med. 2018, 47, 143–158. [Google Scholar]
- Cipriani, N.A. Prognostic Parameters in Differentiated Thyroid Carcinomas. Surg. Pathol. Clin. 2019, 12, 883–900. [Google Scholar] [CrossRef] [PubMed]
- Liu, X.; Bishop, J.; Shan, Y.; Pai, S.; Liu, D.; Murugan, A.K.; Sun, H.; El-Naggar, A.K.; Xiog, M. Highly prevalent TERT promoter mutations in aggressive thyroid cancers. Endocr. Relat. Cancer. 2013, 20, 603–610. [Google Scholar] [CrossRef] [Green Version]
- Vinagre, J.; Almeida, A.; Populo, H.; Batista, R.; Lyra, J.; Pinto, V.; Coelho, R.; Celestino, R.; Prazeres, H.; Lima, L.; et al. Frequency of TERT promoter mutations in human cancers. Nat. Commun. 2013, 4, 2185. [Google Scholar] [CrossRef] [Green Version]
- Landa, I.; Ganly, I.; Chan, T.A.; Mitsutake, N.; Matsuse, M.; Ibrahimpasic, T.; Ghossein, R.A.; Fagin, J.A. Frequent somatic TERT promoter mutations in thyroid cancer: Higher prevalence in advanced forms of the disease. J. Clin. Endocrinol. Metab. 2013, 98, E1562–E15626. [Google Scholar] [CrossRef] [Green Version]
- Cancer Genome Atlas Research, N. Integrated genomic characterization of papillary thyroid carcinoma. Cell 2014, 159, 676–690. [Google Scholar]
- Killela, P.J.; Reitman, Z.J.; Jiao, Y.; Bettegowda, C.; Agrawal, N.; Diaz, L.A., Jr.; Friedman, A.H.; Friedman, H.; Gallia, G.L.; Giovanella, B.C.; et al. TERT promoter mutations occur frequently in gliomas and a subset of tumors derived from cells with low rates of self-renewal. Proc. Natl. Acad. Sci. USA 2013, 110, 6021–6026. [Google Scholar] [CrossRef] [Green Version]
- Kim, T.H.; Kim, Y.E.; Ahn, S.; Kim, J.Y.; Ki, C.S.; Oh, Y.L.; Kim, K.; You, J.W.; Park, W.-Y.; Choe, J.-H.; et al. TERT promoter mutations and long-term survival in patients with thyroid cancer. Endocr Relat Cancer. 2016, 23, 813–823. [Google Scholar] [CrossRef] [Green Version]
- Xing, M.; Liu, R.; Liu, X.; Murugan, A.K.; Zhu, G.; Zeiger, M.A.; Pai, S.; Bishop, J. BRAF V600E and TERT promoter mutations cooperatively identify the most aggressive papillary thyroid cancer with highest recurrence. J. Clin. Oncolog. Off. J. Am. Soc. Clin. Oncol. 2014, 32, 2718–2726. [Google Scholar] [CrossRef] [Green Version]
- Liu, X.; Qu, S.; Liu, R.; Sheng, C.; Shi, X.; Zhu, G.; Murugan, A.K.; Guan, H.; Yu, H.; Wang, Y.; et al. TERT promoter mutations and their association with BRAF V600E mutation and aggressive clinicopathological characteristics of thyroid cancer. J. Clin. Endocrinol. Metab. 2014, 99, E1130–E1136. [Google Scholar] [CrossRef] [Green Version]
- Baxi, V.; Edwards, R.; Montalto, M.; Saha, S. Digital pathology and artificial intelligence in translational medicine and clinical practice. Mod. Pathol. 2022, 35, 23–32. [Google Scholar] [CrossRef] [PubMed]
- Barrera, C.; Velu, P.; Bera, K.; Wang, X.; Prasanna, P.; Khunger, M.; Khunger, A.; Velcheti, V.; Romero, E.; Madabhushi, A. Computer-extracted features relating to spatial arrangement of tumor infiltrating lymphocytes to predict response to nivolumab in non-small cell lung cancer (NSCLC). J. Clin. Oncol. 2018, 36 (Suppl. 15), 12115. [Google Scholar] [CrossRef]
- Kather, J.N.; Heij, L.R.; Grabsch, H.I.; Loeffler, C.; Echle, A.; Muti, H.S.; Krause, J.; Niehues, J.M.; Sommer, K.A.J.; Bankhead, P.; et al. Pan-cancer image-based detection of clinically actionable genetic alterations. Nat. Cancer. 2020, 1, 789–799. [Google Scholar] [CrossRef]
- Fu, Y.; Jung, A.W.; Torne, R.V.; Gonzalez, S.; Vöhringer, H.; Shmatko, A.; Yates, L.R.; Jimenez-Linan, M.; Moore, L.; Gerstung, M. Pan-cancer computational histopathology reveals mutations, tumor composition and prognosis. Nat. Cancer 2020, 1, 800–810. [Google Scholar] [CrossRef] [PubMed]
- Chen, M.; Zhang, B.; Topatana, W.; Cao, J.; Zhu, H.; Juengpanich, S.; Mao, Q.; Yu, H.; Cai, X. Classification and mutation prediction based on histopathology H&E images in liver cancer using deep learning. NPJ Precis Oncol. 2020, 4, 14. [Google Scholar]
- Kather, J.N.; Pearson, A.T.; Halama, N.; Jäger, D.; Krause, J.; Loosen, S.H.; Marx, A.; Boor, P.; Tacke, F.; Neumann, U.P.; et al. Deep learning can predict microsatellite instability directly from histology in gastrointestinal cancer. Nat. Med. 2019, 25, 1054–1056. [Google Scholar] [CrossRef] [PubMed]
- Coudray, N.; Ocampo, P.S.; Sakellaropoulos, T.; Narula, N.; Snuderl, M.; Fenyö, D.; Moreira, A.L.; Razavian, N.; Tsirigos, A. Classification and mutation prediction from non-small cell lung cancer histopathology images using deep learning. Nat. Med. 2018, 24, 1559–1567. [Google Scholar] [CrossRef]
- Schirris, Y.; Gavves, E.; Nederlof, I.; Horlings, H.M.; Teuwen, J. DeepSMILE: Contrastive self-supervised pre-training benefits MSI and HRD classification directly from H&E whole-slide images in colorectal and breast cancer. Med. Image Anal. 2022, 79, 102464. [Google Scholar]
- Lu, M.Y.; Williamson, D.F.K.; Chen, T.Y.; Chen, R.J.; Barbieri, M.; Mahmood, F. Data-efficient and weakly supervised computational pathology on whole-slide images. Nat. Biomed. Eng. 2021, 5, 555–570. [Google Scholar] [CrossRef]
- Lazard, T.; Bataillon, G.; Naylor, P.; Popova, T.; Bidard, F.-C.; Stoppa-Lyonnet, D.; Stern, M.-H.; Decencière, E.; Walter, T.; Vincent-Salomon, A. Deep Learning identifies new morphological patterns of Homologous Recombination Deficiency in luminal breast cancers from whole slide images. Cell Rep. Med. 2022, 3, 100872. [Google Scholar] [CrossRef]
- Tellez, D.; Litjens, G.; Bándi, P.; Bulten, W.; Bokhorst, J.M.; Ciompi, F.; Van Der Laak, J. Quantifying the effects of data augmentation and stain color normalization in convolutional neural networks for computational pathology. Med. Image Anal. 2019, 58, 101544. [Google Scholar] [CrossRef] [Green Version]
- Roy, S.; Kumar Jain, A.; Lal, S.; Kini, J. A study about color normalization methods for histopathology images. Micron 2018, 114, 42–61. [Google Scholar] [CrossRef]
- Campanella, G.; Hanna, M.G.; Geneslaw, L.; Miraflor, A.; Werneck Krauss Silva, V.; Busam, K.J.; Brogi, E.; Reuter, V.E.; Klimstra, D.S.; Fuches, T.J. Clinical-grade computational pathology using weakly supervised deep learning on whole slide images. Nat. Med. 2019, 25, 1301–1309. [Google Scholar] [CrossRef] [PubMed]
- Shi, B.; Bai, X.; Yao, C. An End-to-End Trainable Neural Network for Image-Based Sequence Recognition and Its Application to Scene Text Recognition. IEEE Trans. Pattern Anal. Mach. Intell. 2017, 39, 2298–2304. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Liu, T.; Wang, N.; Cao, J.; Sofiadis, A.; Dinets, A.; Zedenius, J.; Laesson, C.; Xu, D. The age- and shorter telomere-dependent TERT promoter mutation in follicular thyroid cell-derived carcinomas. Oncogene 2014, 33, 4978–4984. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Johnson, J.M.; Khoshgoftaar, T.M. Survey on deep learning with class imbalance. J. Big Data 2019, 6, 27. [Google Scholar] [CrossRef] [Green Version]
- Baloch, Z.W.; Asa, S.L.; Barletta, J.A.; Ghossein, R.A.; Juhlin, C.C.; Jung, C.K.; LiVolsi, V.A.; Papotti, M.G.; Sobrinho–Simões, M.; Tillini, G.; et al. Overview of the 2022 WHO Classification of Thyroid Neoplasms. Endocr. Pathol. 2022, 33, 27–63. [Google Scholar] [CrossRef]
- Huang, G.; Liu, Z.; Maaten, L.V.D.; Weinberger, K.Q. Densely Connected Convolutional Networks. In Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, 21–26 July 2017; pp. 4700–4708. [Google Scholar]
- He, K.; Zhang, X.; Ren, S.; Sun, J. Deep Residual Learning for Image Recognition. In Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 27–30 June 2016. [Google Scholar]
- Tan, M.; Le, Q. EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks. In Proceedings of the 36th International Conference on Machine Learning, Long Beach, CA, USA, 9–15 June 2019; pp. 6105–6114. [Google Scholar]
- Wang, X.; Cheng, W.; Liu, C.; Li, J. Tall cell variant of papillary thyroid carcinoma: Current evidence on clinicopathologic features and molecular biology. Oncotarget 2016, 7, 40792. [Google Scholar] [CrossRef] [Green Version]
- Yang, J.; Gong, Y.; Yan, S.; Chen, H.; Qin, S.; Gong, R. Association between TERT promoter mutations and clinical behaviors in differentiated thyroid carcinoma: A systematic review and meta-analysis. Endocrine 2020, 67, 44–57. [Google Scholar] [CrossRef] [Green Version]
- Kim, H.S.; Kwon, M.J.; Song, J.H.; Kim, E.S.; Kim, H.Y.; Min, K.W. Clinical implications of TERT promoter mutation on IDH mutation and MGMT promoter methylation in diffuse gliomas. Pathol. Res. Pract. 2018, 214, 881–888. [Google Scholar] [CrossRef]
- Lee, H.; Lee, B.; Kim, D.G.; Cho, Y.A.; Kim, J.S.; Suh, Y.L. Detection of TERT Promoter Mutations Using Targeted Next-Generation Sequencing: Overcoming GC Bias through Trial and Error. Cancer Res. Treat. 2022, 54, 75–83. [Google Scholar] [CrossRef]
- McKelvey, B.A.; Zeiger, M.A.; Umbricht, C.B. Characterization of TERT and BRAF copy number variation in papillary thyroid carcinoma: An analysis of the cancer genome atlas study. Genes Chromosomes Cancer 2021, 60, 403–409. [Google Scholar] [CrossRef]
- Tanaka, A.; Matsuse, M.; Saenko, V.; Nakao, T.; Yamanouchi, K.; Sakimura, C.; Yano, H.; Nishihsrs, E.; Hirokawa, M.; Suzuki, K.; et al. TERT mRNA Expression as a Novel Prognostic Marker in Papillary Thyroid Carcinomas. Thyroid 2019, 29, 1105–1114. [Google Scholar] [CrossRef]
- Paulsson, J.O.; Mu, N.; Shabo, I.; Wang, N.; Zedenius, J.; Larsson, C.; Juhlin, C.C. TERT aberrancies: A screening tool for malignancy in follicular thyroid tumours. Endocr. Relat. Cancer 2018, 25, 723–733. [Google Scholar] [CrossRef]
- Dolezal, J.M.; Trzcinska, A.; Liao, C.-Y.; Kochanny, S.; Blair, E.; Agrawal, N.; Keutegen, X.M.; Angelos, P.; Cipriani, N.A.; Pearson, A.T. Deep learning prediction of BRAF-RAS gene expression signature identifies noninvasive follicular thyroid neoplasms with papillary-like nuclear features. Mod. Pathol. 2021, 34, 862–874. [Google Scholar] [CrossRef]
- Tsou, P.; Wu, C.J. Mapping Driver Mutations to Histopathological Subtypes in Papillary Thyroid Carcinoma: Applying a Deep Convolutional Neural Network. J. Clin. Med. 2019, 8, 1675. [Google Scholar] [CrossRef] [Green Version]
- Affinito, O.; Orlandella, F.M.; Luciano, N.; Salvatore, M.; Salvatore, G.; Franzese, M. Evolution of intra-tumoral heterogeneity across different pathological stages in papillary thyroid carcinoma. Cancer Cell Int. 2022, 22, 263. [Google Scholar] [CrossRef]
(a) Tumor and Non-Tumor Patches | (b) TERT ROI for Negative and Positive Cases | |||||||
---|---|---|---|---|---|---|---|---|
Training | Validation | Training | Validation | |||||
Normal | Tumor | Normal | Tumor | Negative | Positive | Negative | Positive | |
CV Set 1 | 26794 | 26417 | 10099 | 5699 | 145 | 225 | 45 | 83 |
CV Set 2 | 24963 | 27675 | 11930 | 4441 | 149 | 239 | 41 | 69 |
CV Set 3 | 28618 | 24019 | 8275 | 8097 | 143 | 236 | 47 | 72 |
CV Set 4 | 34142 | 20825 | 2751 | 11291 | 168 | 246 | 22 | 62 |
CV Set 5 | 33055 | 29528 | 3838 | 2588 | 155 | 286 | 35 | 22 |
(a) CRNN with Normal Transform | (b) CRNN with HSV-Strong | |||||
---|---|---|---|---|---|---|
Precision | Recall | f1-Score | Precision | Recall | f1-Score | |
Negative | 0.93 (±0.13) | 0.84 (±0.19) | 0.87 (±0.12) | 0.97 (±0.03) | 0.89 (±0.18) | 0.92 (±0.11) |
Positive | 0.93 (±0.09) | 0.96 (±0.08) | 0.94 (±0.07) | 0.95 (±0.09) | 0.98 (±0.02) | 0.96 (±0.04) |
Accuracy | 0.92 (±0.08) | 0.95 (±0.06) | ||||
AUC score | 0.90 (±0.09) | 0.94 (±0.08) |
Methods | Sensitivity | Specificity |
---|---|---|
DenseNet161(Norm) + CRNN(Norm) | 0.76 (±0.43) | 0.23 (±0.18) |
DenseNet161(HSV-strong) + CRNN(Norm) | 0.96 (±0.12) | 0.55 (±0.32) |
DenseNet161(HSV-strong) + CRNN(HSV-strong) | 0.99 (±0.00) | 0.60 (±0.31) |
VGG16(Norm) + CRNN(Norm) | 0.78 (±0.34) | 0.33 (±0.29) |
VGG16(HED-strong) + CRNN(Norm) | 0.89 (±0.28) | 0.37 (±0.31) |
VGG16(HED-strong) + CRNN(HSV-strong) | 0.93 (±0.26) | 0.50 (±0.31) |
EfficientNet_b4(Norm) + CRNN(Norm) | 0.92 (±0.22) | 0.51 (±0.30) |
EfficientNet_b4(HSV-light) + CRNN(Norm) | 0.95 (±0.12) | 0.50 (±0.34) |
EfficientNet_b4(HSV-light) + CRNN(HSV-strong) | 0.98 (±0.05) | 0.59 (±0.26) |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Kim, J.; Ko, S.; Kim, M.; Park, N.J.-Y.; Han, H.; Cho, J.; Park, J.Y. Deep Learning Prediction of TERT Promoter Mutation Status in Thyroid Cancer Using Histologic Images. Medicina 2023, 59, 536. https://doi.org/10.3390/medicina59030536
Kim J, Ko S, Kim M, Park NJ-Y, Han H, Cho J, Park JY. Deep Learning Prediction of TERT Promoter Mutation Status in Thyroid Cancer Using Histologic Images. Medicina. 2023; 59(3):536. https://doi.org/10.3390/medicina59030536
Chicago/Turabian StyleKim, Jinhee, Seokhwan Ko, Moonsik Kim, Nora Jee-Young Park, Hyungsoo Han, Junghwan Cho, and Ji Young Park. 2023. "Deep Learning Prediction of TERT Promoter Mutation Status in Thyroid Cancer Using Histologic Images" Medicina 59, no. 3: 536. https://doi.org/10.3390/medicina59030536
APA StyleKim, J., Ko, S., Kim, M., Park, N. J.-Y., Han, H., Cho, J., & Park, J. Y. (2023). Deep Learning Prediction of TERT Promoter Mutation Status in Thyroid Cancer Using Histologic Images. Medicina, 59(3), 536. https://doi.org/10.3390/medicina59030536